Your premier information resource for lone atrial fibrillation

Number 36
4th Year

Download Acrobat Reader
View in .pdf Format
Download Acrobat Reader


The imminent arrival on the drug scene of the new anticoagulant ximelagatran (Exanta) has caused quite a stir in the warfarin camp. After relying for years on the results of strictly controlled clinical trials to extol the virtues of warfarin (Coumadin), medical researchers have finally carried out extensive studies to determine just how effective warfarin is when used under actual "real world" conditions. We discuss the results of these trials in this issue and take a close look at the trial results for ximelagatran. The studies do not change the conclusion that warfarin, or ximelagatran for that matter, does not lower the overall stroke risk (ischemic and hemorrhagic) among lone afibbers with no risk factors for stroke. However, both drugs are generally beneficial for afibbers with heart failure and for afibbers who have experienced a prior stroke.

Also in this issue we report on the results of the latest LAF survey (LAFS-6). I found no evidence that migraine headaches or candida overgrowth are associated neither with the presence of lone atrial fibrillation nor with the severity of episodes. Finally, there is merging evidence that both vitamin B6 and vitamin C are important players in stroke prevention.

Just a reminder - if you haven't already done so, don't forget to get your copy of my book "Lone Atrial Fibrillation: Towards A Cure" at it provides a wealth of information on dealing with LAF.

Yours in sinus rhythm,

Hans Larsen


Vitamin B6 protects against stroke

BOSTON, MASSACHUSETTS. A few years ago it was postulated that high blood levels of the amino acid homocysteine are associated with a significantly increased risk of atherosclerosis and ischemic stroke. It was known that supplementation with folic acid effectively reduces homocysteine levels; it was also known that folic acid is highly effective in preventing neural tube defects in babies. Based on these findings the US government instituted cereal grain folic acid fortification in 1998.

Researchers at the Harvard Medical School and the Massachusetts General Hospital have now re-examined the relationship between vitamin status (folic acid, vitamin B6, and vitamin B12) and stroke risk in the years following the introduction of general folic acid fortification. Their study included 171 patients who had suffered an ischemic stroke (38% cardioembolic and 32% large artery), 9 patients who had experienced a TIA caused by large artery disease, and 140 matched controls. The stroke patients had blood samples drawn for analysis as soon as possible after admission to hospital, but in no case more than 5 days after the event.

The researchers discovered, much to their surprise, that there was no difference in total homocysteine levels between stroke/TIA patients and controls. Both had an average level of 10.8 micromol/L, which is not considered to be significantly elevated. There were, however, significant differences in other measured variables. Stroke/TIA patients were more likely to have hypertension, diabetes and atrial fibrillation; they were also more likely to be smokers and to have lower blood levels of vitamin B6, vitamin B12 and folic acid. However, when all these risk factors were included in a multivariate analysis it became apparent that vitamin B6 level was the major determinant of stroke risk. Participants with a plasma level of PLP (pyridoxal 5'-phosphate, the major metabolite of vitamin B6) of more than 80.1 nmol/L had a 90% lower risk of stroke or TIA than did participants with a PLP level of less than 20 nmol/L. The researchers point out that vitamin B6 is believed to have powerful antiinflammatory properties as evidenced by the strong inverse correlation between C-reactive protein and PLP level.
Kelly, Peter J., et al. Low vitamin B6 but not homocysteine is associated with increased risk of stroke and transient ischemic attack in the era of folic acid grain fortification. Stroke, Vol. 34, June 2003, pp. e51-e54

Editor's comment: These new findings confirm earlier ones by Italian scientists to the effect that high blood levels of PLP (vitamin B6) are associated with a significantly lower risk of deep vein thrombosis[1]. The Italian researchers pointed out that vitamin B6 supplementation has been found to inhibit platelet aggregation and prolong bleeding time as well as protecting against heart attacks and atherosclerosis. The 90% relative reduction in stroke risk among people with high PLP levels is very significant and compares extremely favourably with the oft-quoted relative risk reduction afforded by warfarin (64%) and aspirin (25%). Clearly, ensuring adequate blood levels of PLP is a must for all afibbers. Vitamin B6 is converted to its active metabolite PLP in the liver and there is some evidence that the liver can only handle about 50 mg of vitamin B6 at a time. So taking one or two daily doses of 50 mg of vitamin B6, in conjunction with other B vitamins, is probably safe and sufficient. If a higher level of protection is desired supplementation with PLP directly may be the answer.
[1] Cattaneo, M., et al. Low plasma levels of vitamin B6 are independently associated with a heightened risk of deep-vein thrombosis. Circulation, Vol. 105, November 13, 2001, pp. 2442-46

Antiarrhythmic properties of fish oils

CHIETI, ITALY. Several large clinical trials have confirmed the ability of fish oils to prevent sudden cardiac death in both presumably healthy subjects as well as in patients having suffered a heart attack (myocardial infarction). Considering that sudden cardiac death, largely caused by ventricular fibrillation, accounts for somewhere between 250,000 and 300,000 deaths every year in the US alone, it is clearly highly significant that a diet rich in oily fish or fish oil supplements may reduce the incidence of sudden cardiac death by up to 45%.

Researchers at the Universities of Chieti and Pisa recently published a review of the current "state-of-the-art" in regard to fish oils and arrhythmias. Highlights are:

  • Supplementation with fish oils shows its beneficial effect within a few weeks.
  • It is unlikely that the biological effects of fish oils would vary depending on source (oily fish or fish oil supplement).
  • Animal experiments have shown that fish oils act on individual myocytes (heart cells) to inhibit the excitatory Na+ current, stabilize the inactivated state of the Na channel, and prolong the effective refractory period of the cardiac cycle. The L-type Ca++ current is also inhibited by fish oils and the outward flow of K+ is reduced. All effects which would reduce the tendency to arrhythmia either by decreasing automaticity or by interfering with re- entry circuits.
  • Two small trials have shown a reduction in PVCs (premature ventricular complexes) with fish oil supplementation. In one of these trials 34 participants with frequent PVCs, but no life-threatening arrhythmias were given 2.4 grams/day of fish oils while the control group was given sunflower seed oil which is rich in linoleic acid (an omega-6 fatty acid). PVCs decreased by 48% in the fish oil group as compared to 25% in the sunflower seed group.
  • Fish oils have been shown to decrease heart rate variability and there is some suggestion that they may also reduce sympathetic and increase parasympathetic (vagal) activity in the autonomic nervous system.
  • Prostaglandins and thromboxane A2, produced from arachidonic acid, are mostly proarrhythmic so a high intake of omega-6 fatty acids may be detrimental.

Although most research, so far, has focused on the effect of fish oil on life-threatening ventricular arrhythmias it is likely than many of the findings may also be applicable to atrial fibrillation.
De Caterina, Raffaele, et al. Antiarrhythmic effects of omega-3 fatty acids: from epidemiology to bedside. American Heart Journal, Vol. 146, September 2003, pp. 420-30

Low vitamin C levels associated with increased stroke risk

KUOPIO, FINLAND. Several studies have concluded that low blood levels of vitamin C are associated with an increased risk of dying from a stroke. Finnish researchers now report that low vitamin C levels are also associated with an increased risk of actually having a stroke whether ischemic (caused by a blood clot) or hemorrhagic (caused by a burst blood vessel). Their study involved 2419 randomly selected middle-aged men (42 to 60 years of age) with no history of stroke at the baseline examination. The men were followed for an average of 10.4 years during which time 96 ischemic and 24 hemorrhagic strokes were documented. This corresponds to a total stroke incidence of 0.5% per year.

After adjusting for age, month of examination (vitamin C levels tend to vary with the seasons), body mass index, systolic blood pressure, smoking, alcohol consumption, total cholesterol level, and presence of diabetes or exercise-induced angina, the researchers observed that men with a plasma vitamin C level below 28.4 micromol/L had twice the risk of having a stroke when compared to men with a level above 65 micromol/L. The association was particularly pronounced among hypertensive men where low vitamin C levels were associated with a 2.6 times higher risk and among overweight men where low levels were associated with a 2.7-fold risk increase. The researchers have also observed a significant association between low vitamin C levels and elevated blood pressure (hypertension). They conclude that a low vitamin C level is an independent risk factor for both ischemic and hemorrhagic stroke, especially among hypertensive and overweight men. They call for clinical trials to test the efficacy of vitamin C supplements in the prevention of strokes among hypertensive and overweight (BMI greater than 25 kg/sq m) and obese men.
Kurl, S., et al. Plasma vitamin C modifies the association between hypertension and risk of stroke. Stroke, Vol. 33, June 2002, pp. 1568-73

Editor's comment: Clearly vitamin C is an important component of a comprehensive stroke prevention program. It is particularly important for hypertensive or overweight men and women (an earlier study revealed a similar association between low vitamin C levels and increased stroke risk among women). The 50% relative risk reduction associated with high vitamin C levels compares favourably to the relative reductions quoted for warfarin (64%) and aspirin (25%). Tissue saturation with vitamin C (about 70 micromol/L in plasma) can be obtained by supplementing with 300-500 mg of vitamin C three times a day.

Statin drugs may help prevent atrial fibrillation

HONG KONG, CHINA. Statin drugs such as simvastatin (Zocor) and atorvastatin (Lipitor) are widely used to reduce cholesterol levels. There is also evidence that they are useful in reducing systemic inflammation. Medical researchers at the University of Hong Kong now report that statins may also be effective in preventing the recurrence of atrial fibrillation. Their study involved 62 patients with lone persistent AF who had undergone successful electrical cardioversion between July 1998 and December 1999. The average age of the patients was 61 years, 74% were men and 16% (10 patients) were taking statin drugs to control their cholesterol levels.

After two years of follow-up the rate of recurrence of afib was 40% in the statin group and 84% in the control group for a relative risk reduction of 69%. The researchers believe that the anti-inflammatory properties of statins are involved in the observed risk reduction, but also point out that, "statins may also exhibit direct antiarrhythmic effects by modulating the fatty acid composition and physiochemical properties of cell membranes, with resultant alleviations in transmembrane ion channel properties".

A similar study carried out by American researchers supports the idea that statins may help prevent AF. This study involved 449 patients with coronary artery disease, but no AF at baseline. During a 5-year follow-up period 52 patients (12%) developed AF. The researchers noted that statin therapy was associated with a 50- 60% lower risk of developing AF and conclude that statin therapy in patients with chronic, stable coronary artery disease appear to be protective against AF. They point out that the mechanism underlying the protective effect is unknown, but appears to be independent of the reduction in cholesterol levels.
Siu, Chung-Wah, et al. Prevention of atrial fibrillation recurrence by statin therapy in patients with lone atrial fibrillation after successful cardioversion. American Journal of Cardiology, Vol. 92, December 1, 2003, pp. 1343-45
Young-Xu, Y, et al. Usefulness of statin drugs in protecting against atrial fibrillation in patients with coronary artery disease. American Journal of Cardiology, Vol. 92, December 15, 2003, pp. 1379-83

Editor's comment: This is a very interesting finding and confirms that reducing inflammation and modifying cell membranes may be important aspects of afib prevention. Statin drugs, unfortunately, have several undesirable potential side effects including memory loss and muscle disease. These drugs also severely deplete coenzyme Q10 levels. If inflammation reduction and membrane modification are indeed the "name of the game" then it is quite possible that high doses of fish oils (5 grams/day) may have similar effects.

On-demand approach found effective

PIACENZA, ITALY. Many afibbers have successfully used the "on-demand" method to terminate an episode. This approach involves swallowing 200 mg of crushed flecainide (Tambocor) or 300 mg of crushed propafenone (Rythmol) with warm water as soon as possible (preferably within 5-10 minutes) after the onset of an episode. Italian researchers have now evaluated the approach in a clinical trial and have found it to be safe and effective.

Their study involved 212 afibbers admitted to hospital for a first AF episode (65% male, 65 years average age). Oral propafenone was administered according to the weight of the patient � 600 mg if the patient's weight was 70 kg or more, otherwise 450 mg. Patients were excluded from the trial if they had been diagnosed with heart failure, recent heart attack, unstable angina, electrocardiographic evidence (present or past) of ventricular preexcitation, complete bundle branch block, sick sinus syndrome, hypokalemia (potassium level < 3.5 mEq/L), renal or hepatic insufficiency, severe hypoxia (arterial partial pressure of oxygen < 55 mm Hg), severe metabolic disturbances or thyroid dysfunction. The patients were kept in the supine position after the drug was administered.

The average time elapsed from episode onset to drug administration was 8 hours and average conversion time was 3.5 hours. No ventricular proarrhythmic effects were observed. During follow-up 65 patients were readmitted due to a second episode. They were again given propafenone orally, but this time the elapsed time to drug administration was shorter at an average 3.2 hours and the average time to conversion was also shorter at 2.4 hours. The conversion efficacy was 92%.

The researchers conclude that the "on-demand" or "pill-in-the-pocket" approach may be appropriate for home use in patients who can correctly self-diagnose the symptoms of recurrent afib and can self-manage the therapy (i.e., tablet intake and bed rest until conversion to sinus rhythm).
Capucci, A, et al. Reproducible efficacy of loading oral propafenone in restoring sinus rhythm in patients with paroxysmal atrial fibrillation. American Journal of Cardiology, Vol. 92, December 1, 2003, pp. 1345-47

Editor's comment: These findings are indeed good news. Although several afibbers have been using the on-demand approach for over a year now with good success it is always comforting to know that the approach has been found safe and effective in a clinical trial. The dosage of propafenone used in the Italian study (600 mg or 450 mg) is somewhat higher than the 300 mg used by most afibbers. However, it is quite likely that the 300 mg is equally effective if taken early, crushed, and with warm water. It is interesting that patients were kept supine until conversion. I have evaluated the effect of lying down after taking the propafenone rather than just carrying on as usual. It does seem to work faster in the supine position, perhaps because this position would give the heart rate-slowing parasympathetic system a bit of a boost.

Survey Results � LAFS-6

The 6th lone atrial fibrillation survey (LAFS-6), our first web-based survey, was designed to determine a possible association between LAF and migraine headaches and between LAF and the presence of candida overgrowth (yeast infection). A total of 65 lone afibbers without heart disease participated in the survey. Twenty-three per cent of respondents were women, a proportion similar to that observed in previous surveys. The average age of respondents was 55 years (range of 31 to 84 years). The average age at diagnosis was 49 years (range of 20 to 77 years) and the average number of years of having lived with afib was 6 years. There were no significant differences between men and women in regard to age or years of afib.

The majority (87%) of respondents had the paroxysmal (intermittent) kind of LAF; 5% had persistent LAF, and 8% the permanent (chronic) variety. Of the paroxysmal afibbers, 13% had the adrenergic type, 45% the mixed, and 42% were vagal.

Episode frequency was distributed as follows:

  • No episodes in 6-month period ������ 7%
  • 1-3 episodes in 6-month period ������ 21%
  • 4-6 episodes in 6-month period ������ 12%
  • 7-10 episodes in 6-month period �����.. 16%
  • 11-20 episodes in 6-month period �����. 16%
  • More than 20 episodes in 6-month period �� 28%

Episode duration was distributed as follows:

  • No episodes �������� 7%
  • Episodes lasting 0.1-1 hrs �... 16%
  • Episodes lasting 1-2 hrs ��� 7%
  • Episodes lasting 3-6 hrs ��� 14%
  • Episodes lasting 7-12 hrs ��.. 22%
  • Episodes lasting 13-19 hrs ��. 10%
  • Episodes lasting 20-29 hrs ��.. 3%
  • Episodes lasting 30-48 hrs ��. 10%
  • More than 48 hrs ������.. 11%

Twelve respondents (18.5%) had undergone ablation therapy. Six of the procedures (50%) had been successful. This success rate is somewhat low, partly due to the fact that 2 of the so far unsuccessful ablations were performed quite recently. Of some concern is the finding that afibbers who had reported an unsuccessful ablation prior to 2003 were likely to be experiencing more and longer-lasting episodes than did afibbers who had not undergone ablation. These findings confirm observations from earlier surveys and clearly need further investigation.

Fifty-two per cent of all respondents were taking pharmaceutical drugs on a regular basis to manage their LAF, while 22% were using the on-demand approach (200 mg flecainide or 300 mg propafenone crushed and swallowed with warm water immediately at onset of an episode). There was no indication that drug use, regular or on-demand, had any effect on the average episode frequency or duration. There was also no indication that taking statin drugs or fish oils affected episode frequency or duration. However, there was a slight, statistically non-significant (p=0.06) trend for women to be more likely to take drugs on a regular basis.

Fourteen of 65 respondents had experienced a migraine headache in the past and eight were still experiencing them (1-3 in 6-month period). Thus the current prevalence of migraines among the 65 respondents is 12.3%, not significantly different from the 10.3% experienced among the general US population. There was no correlation between the number of migraine attacks and the frequency or duration of afib episodes. Only one out of 11 had ever noticed an association between migraines and afib episodes. There was no indication that afib type (adrenergic, mixed, vagal, permanent) was related to the prevalence of migraines. Thus it is unlikely that there is an association between migraine headaches and LAF.

Forty-one of 65 respondents had done the saliva test for candida overgrowth ( and 41% had tested positive, 49% negative, and 10% were unsure. Thirteen or 20% of the 65 respondents had been officially diagnosed with candida at one point or other and 10 (77%) of these were still battling the disorder when completing the survey. The most common location for candida overgrowth was the colon (46%) followed by the genital area (38%), and mouth (16%). Two respondents had overgrowth in more than one location. Women were more likely to have been diagnosed with candida than were men; however, there was no gender- related difference according to the saliva test. This could indicate that candida overgrowth is under-diagnosed in men.

Of the 13 respondents who had been diagnosed with candida 10 had taken the saliva test. Seven (70%) of these tested positive, 2 tested negative, and 1 was not sure. Of the 10 respondents currently battling candida 8 had taken the saliva test and 7 (88%) tested positive. Thus it would appear that the saliva test is a reasonably accurate indication of the presence of candida overgrowth. It would also appear that candida is difficult to overcome on a permanent basis.

There was no association between the presence of candida and episode frequency or duration. Two out of 10 respondents had observed an association between candida episodes and afib episodes.

Seven out of 65 respondents (11%) were taking statin drugs for cholesterol problems. There was no indication that this affected episode frequency or duration.

Fish oils
Forty-nine per cent of all respondents were supplementing with fish oils, but there was no indication that fish oil supplementation correlated with episode frequency or duration. Women were slightly more likely to be taking fish oils than were men.

In conclusion, there would appear to be no association between migraine headaches and LAF nor between candida overgrowth and LAF.


Ximelagatran � A Promising New Anticoagulant

Ischemic stroke (cerebrovascular event), heart attack (myocardial infarction), and pulmonary embolism are major causes of death and disabling disease. Ischemic stroke is caused by an embolism, that is, the lodging of a blood clot (thrombus) or a fragment of atherosclerotic plaque in a small artery of the brain, resulting in the cutting off of blood supply to part of the brain. Pulmonary embolism is caused by a thrombus formed in the veins (usually in the legs) that is carried to the lungs in the blood stream and gets stuck there. Heart attack is caused by the build-up of atherosclerotic plaque or a blood clot in the coronary arteries.

Thrombi can form in both the arterial system (the supply side involving the left side of the heart) and in the venous system (the return side involving the right side of the heart and the lungs). The formation of a blood clot involves platelet activation, platelet aggregation, and blood coagulation. The body has a built-in process for dealing with clotted blood that no longer serves its purpose of preventing excessive bleeding. This process is called fibrinolysis.

Stroke, heart attack, and pulmonary embolism are extremely common and often fatal conditions. It is estimated that about 700,000 strokes (mostly ischemic), 650,000 cases of pulmonary embolism, and 1.5 million heart attacks occur every year in the United States alone[1-3]. Not surprisingly, this state of affairs has spawned the development of several drugs designed to reduce the natural tendency of blood to coagulate when flowing too slowly (in the veins) or when exposed to a damaged blood vessel wall. The most popular drug for preventing platelet activation and aggregation is aspirin and the most popular drug for preventing blood coagulation is warfarin (Coumadin). Heparin is also effective in preventing coagulation, but is much less used as it needs to be given by injection.

Several trials have shown that aspirin is quite effective in protecting against heart attack, particularly a second one. It is less effective in protecting against stroke and for "serious" stroke protection in high-risk individuals warfarin is currently the drug of choice. Warfarin is also more effective in preventing deep vein thrombosis, a condition often leading to pulmonary embolism.


Warfarin was first isolated in the 1920s after farmers noted that their cows often bled to death after eating spoiled sweet clover. It was patented as a powerful rat poison in 1948. By the mid-fifties it was beginning to be used as an anticoagulant in humans. The drug lost in popularity during the 1970s when it was realized that it probably caused as many deaths from bleeding as it prevented deaths from stroke. One of the major problems was the need to adjust the dosage for each individual patient. This, combined with the fairly unreliable test methods used at the time, resulted in a less than sterling experience with the drug. For the last 20 years warfarin has experienced a resurgence, partly because of the development of an improved test for bleeding time (International Normalized Ratio) and partly because of a superb marketing campaign by the manufacturer of Coumadin.

Several studies were carried out in the early 1990s to determine the effectiveness of warfarin in preventing stroke in atrial fibrillation patients. Unfortunately, these studies did not distinguish between AF with underlying heart disease and AF without heart disease (lone atrial fibrillation). One major trial (SPAF II) specifically excluded lone afibbers and all trials included a large proportion of patients not only with heart disease, but also with one or more risk factors for stroke (hypertension, diabetes, heart failure or a prior stroke or TIA). Thus the applicability of the trial data to lone afibbers, and in particular lone afibbers without risk factors, is very much in doubt. The average annual rates of ischemic stroke among all patients included in the five trials were as follows[4]:

Annual incidence of stroke, %
Relative risk reduction
Age less than 65 years
No risk factors
One or more risk factors
Age 65 to 75 years
No risk factors
One or more risk factors
Age over 75 years
No risk factors
One or more risk factors

It is clear that there is no reason to prescribe warfarin for afibbers below the age of 65 years who have no additional risk factors for stroke and indeed this is fully recognized in the current guidelines for the management of atrial fibrillation[5].

The data would, however, indicate that prescribing warfarin for afibbers over the age of 75 years having one or more risk factors might be prudent. The SPAF II trial found that the risk of an ischemic stroke in this age group was 3.6% when on warfarin and 4.8% when on aspirin. However, when looking at the combined total of fatal and disabling ischemic and hemorrhagic strokes there was little difference � 4.6% in the warfarin group and 4.3% in the aspirin group. So again, the wisdom of prescribing warfarin rather than aspirin for older afibbers is by no means clear-cut. It should also be kept in mind that the SPAF II trial was a clinical trial with frequent and accurate monitoring of INR levels. Even though close monitoring would presumably reduce the risk of major bleeding, the SPAF II study found the risk of major internal bleeding in patients over 75 years to be 4.2%, thus largely cancelling out the benefit of ischemic stroke protection[6].

More recently, a group of American researchers evaluated stroke risk among 700 elderly participants (mean age of 75 years) in the Framingham Heart Study and concluded that afibbers with a predicted annual stroke risk of 2% or less (irrespective of age) may not realize additional benefit from warfarin compared with aspirin and their risk of stroke may not exceed the risk of life-threatening bleeding with warfarin. Thus anticoagulation therapy may not be justified in individuals with low predicted rates of stroke[7,8].

A team of American, Canadian, Dutch, and Danish medical researchers looked at the effectiveness of warfarin therapy among participants in six major trials. The annual rate of stroke in the low-risk group (no hypertension, angina or diabetes and no history of stroke or TIA) was 1.5% on warfarin as compared to 1.0% with aspirin and 1.2% in an age and gender matched cohort without afib. The stroke risk in the remaining moderate- to high-risk group was 3.4% per year with warfarin, 4.2% with aspirin, and 1.3% in an age and gender matched cohort without afib or risk factors. The researchers conclude that, irrespective of age, afibbers who satisfy the criteria for low risk can safely take aspirin for stroke prevention and would not benefit from oral anticoagulation. They estimate that about one quarter of all afibbers would fall in the low-risk group[9].

Researchers at Kaiser Permanente in northern California recently concluded that results regarding warfarin efficacy obtained in tightly controlled clinical trials may not necessarily be indicative of what is going on in the "real world". They followed 11,526 patients with nonvalvular atrial fibrillation for an average of 2.2 years (25,341 person years). About half the patients (6,320) were treated with warfarin while the remainder (5,089) took daily aspirin or used no drugs for stroke prevention. The average age of the patients was 71 years with about 40% being over the age of 74 years and about 24% being below the age of 65 years. Most of the participants had one or more risk factors for stroke (previous ischemic stroke [8%], heart failure [28.5%], hypertension [50.1%], diabetes [16.8%], and coronary heart disease [27.7%]). Almost half (43%) of patients were women. This survey population thus has little in common with a representative group of lone afibbers.

During the follow-up period the researchers observed 141 ischemic strokes, 59 hemorrhagic strokes (intracranial bleeding), and 118 major gastrointestinal bleeds in the warfarin group and 231 ischemic strokes, 29 hemorrhagic strokes, and 119 major gastrointestinal bleeds in the aspirin/no drug therapy group. Results are detailed below[10].

Incident Rate, % per year

Ischemic Stroke
Hemorrhagic Stroke*
Major Bleed*
Total Major Incidents
Patients on Warfarin
No risk factors
Prior stroke
Heart failure
Heart disease
Age over 75 years
Total population
Patients on Aspirin or No Drug Therapy
No risk factors
Prior stroke
Heart failure
Heart disease
Age over 75 years
Total population

* Assuming that these rates are independent of ischemic stroke risk factors
A = ischemic + hemorrhagic stroke
B = ischemic + hemorrhagic stroke + major bleeding

Difference in Annual Incident Rate
Warfarin vs. No Warfarin

Ischemic +
Ischemic + Hemorrhagic
Ischemic Stroke
Hemorrhagic Stroke
Stroke + Major Bleed
Net Benefit, % per year
No risk factors
0.22 (51%)
-0.01 (0%)
0.04 (2%)
Prior stroke
4.16 (56%)
4.85 (63%)
4.90 (57%)
1.5 (42%)
1.27 (34%)
1.32 (28%)
0.96 (38%)
0.73 (26%)
0.78 (21%)
Heart failure
2.32 (66%)
2.09 (55%)
2.14 (45%)
Heart disease
1.37 (47%)
1.14 (36%)
1.19 (40%)
Age over 75 years
1.79 (56%)
1.56 (45%)
1.61 (37%)
Total population
0.77 (41%)
0.53 (25%)
0.59 (19%)

The percentages in brackets indicate the relative superiority of warfarin over aspirin/no drug therapy.

In reviewing the above results it should be kept in mind that a hemorrhagic stroke can be just as devastating as an ischemic stroke, so what really matters to the patient is the combined incidence of the two. The incidence of major gastrointestinal hemorrhage (defined as death or hospitalization requiring blood transfusion) was similar in the two groups at about 1% per year. This is no doubt due to the fact that the non-warfarin group included patients taking aspirin. Regular aspirin usage has been associated with a 2.5% annual bleeding risk in other studies[11].

It is clear that warfarin does indeed have a moderate overall beneficial effect on the incidence of stroke (ischemic and hemorrhagic) in patients with non-valvular atrial fibrillation. For every 100 patients treated with warfarin for a year 0.53 strokes are avoided. This corresponds to a relative risk reduction of 25% when compared to patients taking aspirin or using no drug therapy. The benefit of warfarin therapy is substantial for patients having suffered a prior ischemic stroke (4.85 strokes per year avoided or a 63% benefit), but non- existent for afibbers with no risk factors for stroke (zero strokes per year avoided or a 0% benefit). This finding, of course, is in line with numerous previous studies. The benefit of warfarin therapy for afibbers with hypertension is 0.73 strokes avoided per year for a relative benefit of 26%. This is significant, but not impressive.

Overall, it is apparent that 1,000 patients need to be treated with warfarin for a year in order to prevent 5.3 strokes. Treating one patient for a year would cost an estimated $830 (8 lab tests @ $50/test + 4 doctor's visits per year @ $75/visit + drug cost @ $130). Thus treating 1,000 patients would cost $830,000 or $157,000 per annual stroke avoided. The cost per annual stroke avoided in patients with a prior stroke would be $20,000 and that for patients with hypertension $86,000 per stroke per year. The cost would, of course, be astronomical for patients with no risk factors.

The Kaiser Permanente researchers conclude that, "Warfarin is very effective in preventing ischemic stroke in patients with atrial fibrillation in clinical practice while the absolute increase in the risk of intracranial hemorrhage is small".

I find this conclusion hard to reconcile with the actual data presented in the report. I would conclude that warfarin therapy is contraindicated for afibbers without risk factors, is quite effective for patients with heart failure and for those who have suffered a stroke previously, and is marginally effective for afibbers with hypertension. However, the cost of warfarin therapy to the healthcare system is considerable with an estimated annual cost of about $86,000 per stroke avoided among afibbers with hypertension.

A group of 7500 California Medicaid recipients with afib and one or more of the following conditions � hypertension (58%), congestive heart failure (48%), diabetes (34%), prior stroke (17%) or prior heart attack (14%) participated in a recent study to evaluate the effectiveness of warfarin therapy. During follow-up stroke occurred in 514 patients with a rate of 3.4 per 100 person-years in patients treated with warfarin and a rate of 4.1 per 100 person-years for those not on warfarin. This corresponds to an overall absolute risk reduction of 0.7% per year. Bleeding occurred in 302 patients with a rate of 3.0 per 100 person-years in patients treated with warfarin and a rate of 2.2 per 100 person-years for those not on warfarin. This corresponds to an absolute increase in bleeding risk of 0.8% per year. The researchers conclude that, "Warfarin therapy, in clinical practice, has a relatively modest benefit in terms of reducing stroke rates, with the greatest benefit occurring among patients with moderate stroke risk. However, this benefit is somewhat offset by the increased risk of bleeding events"[12].

Unfortunately, quite apart from its limited efficacy, warfarin also has other significant shortcomings:

  • Reliably maintaining an INR within the customary range of 2.0 to 3.0 is still a very dicey proposition. Two recent, tightly-controlled clinical trials found that only 57% and 66% respectively of participants were consistently within the desired range during the trial[13,14]. The number of patients within the range in a much less controlled actual daily practice would be considerably lower. Having an INR above 3.0 increases the risk of internal bleeding while having an INR below 2.0 significantly increases the risk of thrombus formation.

  • Appropriate monitoring of INR levels requires frequent visits to a testing laboratory resulting in considerable inconvenience to the patient and a significant financial burden on the health care system.

  • INR levels are strongly affected by many foods and herbs adding further to the difficulty of maintaining the desired ratio.

  • Warfarin interacts with at least 60 common drugs. Some of the interactions, particularly with acetaminophen (Tylenol, Paracetemol), can be fatal as a combination of warfarin and acetaminophen can raise INR to 6.0 or higher[15-17].

  • Warfarin can cause hemorrhagic stroke, gastrointestinal bleeding, osteoporosis and bone fractures (with long-term use), skin necrosis in some cases involving amputation of breast or penis, and no doubt a host of other less common complications[15,18].

The many shortcomings of warfarin and the relative under-utilization of anticoagulation for prevention of thrombosis and embolism have led to a concerted effort to find an effective replacement. It now appears that the search may have succeeded with the development of the new oral anticoagulant ximelagatran.


Ximelagatran (Exanta) was developed by the Swedish arm of AstraZeneca and by now is probably one of the most carefully tested of all pharmaceutical drugs. Ximelagatran or rather its metabolite, melagatran, prevents blood coagulation by directly inhibiting the final step in the coagulation process � namely, the conversion of fibrinogen to insoluble fibrin by thrombin. Warfarin, on the other hand, works less directly by reducing the blood level of vitamin K-dependent coagulation factors. Early clinical trials of ximelagatran concluded that it has many advantages over warfarin.

  • One size fits all. Two 36 mg tablets of ximelagatran taken daily provide anticoagulation equivalent to or better than that afforded by warfarin. There is no need for periodic monitoring of clotting time.

  • Ximelagatran is not metabolized by the cytochrome P450 enzyme system in the liver so it's effect is not altered by foods, herbs or supplements. There is no need to be concerned about what foods or supplements are safe to take when on ximelagatran.

  • There are no known interactions between ximelagatran and other pharmaceutical drugs.

  • Ximelagatran does increase bleeding risk, but the increase in risk is no greater than what is experienced with warfarin in the 2.0-3.0 INR range.

  • Ximelagatran has no known major adverse effects in the short-term (about 18 months); however, it is not known whether it may have adverse effects in the long-term. A significant elevation in liver enzymes has been noticed in about 5% of patients, but this increase usually disappears with continued use or can be reversed by discontinuing the drug.

It is clear that ximelagatran has many advantages over warfarin, but is it equally effective?

Clinical trials of ximelagatran
Several clinical trials have evaluated the effectiveness of ximelagatran as compared to warfarin. For ethical reasons, ximelagatran has not been evaluated against placebo. One trial involved heart attack patients, two trials involved patients with venous thromboembolism while the five SPORTIF (Stroke Prevention Using Oral Thrombin Inhibitor in Atrial Fibrillation) trials involved patients with atrial fibrillation.

Ximelagatran in heart disease patients
This trial (ESTEEM) involved 1833 patients who had suffered a heart attack. The patients were randomized to placebo or 24 mg, 36 mg, 48 mg or 60 mg of ximelagatran twice daily. All participants also received 160 mg of aspirin daily. During the 6-month duration of the trial 16.3% of the placebo group members suffered another non-fatal heart attack, died or developed severe recurrent ischemia. The corresponding number for the combined ximelagatran groups was 12.7% indicating that adding ximelagatran to a daily aspirin can reduce risk by an absolute 3.6% or a relative reduction of 22% over and above the protection provided by aspirin alone. The incidence of major bleeding was 1.8% in the combined ximelagatran group and 0.9% in the aspirin only (placebo) group. The concentration of the liver enzyme, alanine transaminase, was more than 3 times the upper normal limit in 4% of the participants receiving 24 mg of ximelagatran twice daily as compared to 1% in the placebo group. The underlying mechanism for this increase is under investigation. No major adverse effects were observed during the 6-month trial. The researchers conclude that aspirin plus 24 mg of ximelagatran twice daily provides an absolute additional risk reduction of 4% over aspirin alone, has an acceptable risk of major bleeding, and a manageable rise in liver enzymes. It is of interest to note that the total incidence of internal bleeding (minor and major) was 22% in the combined ximelagatran group versus 13% in the aspirin only group[19].

Ximelagatran in prevention of venous thromboembolism
This Swedish trial involved 1233 patients from 18 countries with confirmed venous thromboembolism who had been treated for 6 months with anticoagulant therapy without reoccurrence of thromboembolism. The participants were randomized to receive placebo or 24 mg of ximelagatran twice daily for an 18-month period after discontinuing warfarin therapy. No coagulation tests were carried out during the trial period. At the end of the period 12 patients (2%) in the ximelagatran group had experienced a new thromboembolism versus 71 patients (12%) in the placebo group � a 10% absolute and an 84% relative risk reduction. The risk of minor or major internal bleeding was 22% and 18% in the ximelagatran and placebo groups respectively. Major hemorrhage accounted for about 1% in each group. The cumulative incidence of elevation of alanine aminotransferase to more than 3 times upper normal level was 6.4% in the ximelagatran group and 1.2% in the placebo group. The researchers conclude that oral ximelagatran is effective in extended prevention of venous thromboembolism[20].

Ximelagatran after knee replacement
Venous thromboembolism occurs in 40 to 84% of patients undergoing total knee replacement if they do not receive anticoagulation therapy. About 7% of patients experience a pulmonary embolism and in about 0.2 to 0.7% of cases the embolus is fatal. Anticoagulation with warfarin reduces the incidence of venous thromboembolism to about 47%.

A group of researchers from Brazil, Canada, Israel, Mexico and the United States has just completed a study to determine the relative efficacy of warfarin and ximelagatran in the prevention of venous thromboembolism after total knee replacement. The 1851 study participants were randomized to receive warfarin (INR = 2.0-3.0) or 24 mg or 36 mg of oral ximelagatran twice daily for 7-12 days following surgery. At the end of the trial 128 patients in the 36 mg ximelagatran group (20%) had either died or developed venous thromboembolism while 168 patients in the warfarin group (28%) had done likewise. Minor or major bleeding occurred in 5.3% of the patients in the 36 mg ximelagatran group and in 4.5% of those in the warfarin group. (NOTE: This after only 7-12 days of therapy). Levels of alanine aminotransferase were not noticeably elevated in either group after this short treatment. The researchers conclude that 36 mg of ximelagatran administered twice daily is superior to warfarin in preventing venous thromboembolism following total knee replacement[21].

Ximelagatran in atrial fibrillation
Following a small preliminary trial (SPORTIF II) to investigate optimal dosing and alleviate safety concern two major trials were undertaken to evaluate the efficacy of ximelagatran in stroke prevention among patients with non-valvular atrial fibrillation[4]. SPORTIF III involved 3407 patients from 23 countries and SPORTIF V involved 3922 patients from 409 locales in North America. All study participants had one or more risk factors for stroke (about 40% had coronary artery disease and about 75% had hypertension) and 89% of participants had persistent rather than paroxysmal afib. Thus results are not directly applicable to paroxysmal afibbers with no risk factors for stroke. The 7300 high-risk patients were randomized to receive either 36 mg of ximelagatran twice daily or warfarin adjusted to an INR of 2.0-3.0. After an average follow-up of about 18 months 91 patients (1.6%) in the combined ximelagatran groups (SPORTIF III and V) had suffered an ischemic or hemorrhagic stroke or had experienced a systemic thromboembolic event. The corresponding figure for the combined warfarin groups was 93 patients (1.6%). Minor and major bleeding events were significantly lower in the ximelagatran groups at 37% versus 47% in the warfarin groups (NOTE: A rate of 37% or 47% of patients experiencing internal bleeding during an 18-month period is still, or ought to be, a major cause for concern). If deaths, strokes and major bleeding events were combined the incidence was 5.2% in the ximelagatran groups compared to 6.2% with warfarin. A significant relative risk reduction of 16%. Elevated liver enzyme levels (alanine aminotransferase) were noted in about 6% of ximelagatran users and in about 1% of warfarin users. The investigators conclude that fixed dose ximelagatran is at least as effective as well-controlled warfarin in preventing stroke and systemic embolism in high-risk patients with atrial fibrillation[14,22,23].

Trials involving close to 10,000 patients have shown that ximelagatran is equivalent or superior to warfarin in preventing stroke and systemic embolism. Bleeding rates are similar, although, in my opinion, unacceptably high if applied to patients at low risk for stroke. Significant elevation of liver enzymes occurs in about 5% of patients on ximelagatran, but appears to be fully reversible. No major adverse effects have been observed in trials lasting up to 20 months, but longer term effects are unknown.

Ximelagatran is superior to warfarin in ease of administration and control. The dosage is fixed at 24 or 36 mg twice daily and no monitoring of coagulation parameters is required. It also has the great advantage of not being affected by different foods, herbs and supplements and not interacting with other commonly used drugs. However, liver enzymes do need to be checked monthly for the first 6 months of therapy.

Ximelagatran has now successfully passed phase III trials and is awaiting approval in Europe, the USA and Canada.

While ximelagatran would likely be advantageous for afibbers with one or more risk factors for stroke, I do not believe it would be any more appropriate for afibbers with no risk factors than is warfarin.


  1. Broderick, JP. Stroke therapy in the year 2025: burden, breakthroughs, and barriers to progress. Stroke, December 11, 2003
  2. Feied, Craig. Pulmonary embolism.
  4. Falk, Rodney H and Podrid, PJ, editors. Atrial Fibrillation: Mechanisms and Management. 2nd edition, 1997, Lippincott- Raven Publishers, Philadelphia, p. 282
  5. ACC/AHA/ESC guidelines for the management of patients with atrial fibrillation: executive summary. Circulation, Vol. 104, October 23, 2001, pp. 2118-50
  6. Warfarin versus aspirin for prevention of thromboembolism in atrial fibrillation: Stoke Prevention in Atrial Fibrillation II Study. The Lancet, Vol. 343, March 19, 1994, pp. 687-91
  7. Wang, Thomas J, et al. A risk score for predicting stroke or death in individuals with new-onset atrial fibrillation in the community. JAMA, Vol. 290, August 27, 2003, pp. 1049-56
  8. Waldo, Albert L. Stroke prevention in atrial fibrillation. JAMA, Vol. 290, August 27, 2003, pp. 1093-95
  9. van Walraven, Carl, et al. A clinical prediction rule to identify patients with atrial fibrillation and a low risk for stroke while taking aspirin. Archives of Internal Medicine, Vol. 163, April 28, 2003, pp. 936-43
  10. Go, AS, et al. Anticoagulation therapy for stroke prevention in atrial fibrillation. JAMA, Vol. 290, November 26, 2003, pp. 2685-92
  11. Derry, S and Loke, YK. Risk of gastrointestinal haemorrhage with long term use of aspirin. British Medical Journal, Vol. 321, November 11, 2000, pp. 1170-71, 1183-87
  12. Warfarin of only modest benefit. Circulation, Vol. 108, No. 17, October 28, 2003, p. IV-757, abstract #3419
  13. Petersen, Palle, et al. Ximelagatran versus warfarin for stroke prevention in patients with nonvalvular atrial fibrillation. Journal of the American College of Cardiology. Vol. 41, May 7, 2003, pp. 1445-51
  14. Stroke prevention with the oral direct thrombin inhibitor ximelagatran compared with warfarin in patients with non-valvular atrial fibrillation (SPORTIF III). The Lancet, Vol. 362, November 22, 2003, pp. 1691-98
  15. Wintrobe's Clinical Hematology, 9th edition, Lea & Febiger, 1993, p. 1530-31
  16. Wells, PS, et al. Interactions of warfarin with drugs and food. Annals of Internal Medicine, Vol. 121, November 1, 1994, pp. 676-83
  17. Hylek, EM, et al. Acetaminophen and other risk factors for excessive warfarin anticoagulation. Journal of the American Medical Association, Vol. 279, March 4, 1998, pp. 657-62, 702-03
  18. Carabballo, PJ, et al. Long-term use of oral anticoagulation and the risk of fracture. Archives of Internal Medicine, Vol. 159, August 9/23, 1999, pp. 1750-56
  19. Wallentin, Lars, et al. Oral ximelagatran for secondary prophylaxis after myocardial infarction: The ESTEEM randomised controlled trial. The Lancet, Vol. 362, September 6, 2003, pp. 789-97
  20. Schulman, Sam, et al. Secondary prevention of venous thromboembolism with the oral direct thrombin inhibitor ximelagatran. New England Journal of Medicine, Vol. 349, October 30, 2003, pp. 1713-21
  21. Francis, Charles W., et al. Comparison of ximelagatran with warfarin for the prevention of venous thromboembolism after total knee replacement. New England Journal of Medicine, Vol. 349, October 30, 2003, pp. 1703-12
  22. Halperin, JL, et al. Ximelagatran compared with warfarin for prevention of thromboembolism in patients with nonvalvular atrial fibrillation: rationale, objectives, and design of a pair of clinical studies and baseline patient characteristics (SPORTIF III and V). American Heart Journal, Vol. 146, September 2003, pp. 431-38
  23. Haperin, JL. SPORTIF V: Stroke prevention using oral thrombin inhibitor in atrial fibrillation.


The AFIB Report is published 10 times a year by Hans R. Larsen MSc ChE
1320 Point Street, Victoria, BC, Canada V8S 1A5
Phone: (250) 384-2524
E-mail: [email protected]
Copyright © 2004 by Hans R. Larsen

The AFIB Report does not provide medical advice. Do not attempt self- diagnosis or self-medication based on our reports. Please consult your health-care provider if you wish to follow up on the information presented.